10+ years in tech: Engineering + AI + platform. hackathon winner. open source contributor. kickball mvp. Benchmarking models https://xitter.twitterapi.xyz/t.co/aMN2ZSpVCy
Remote
Joined December 2025
- Tweets4.2K
- Following764
- Followers878
- Likes19.8K
Pinned Tweet
GPT-6.1 Terra fixed all 20 bugs in my app for about $.06 a run
GPT-6.1 SOL did the same for $.22 and opus 5.5 for $.78
GPT-6 Luna and Haiku 5.5 cost less but missed bugs in every run
No one working on anything serious is using Gemini models.
The adoption numbers are because they shoved Gemini in every google search
4
3
238
im breaking down where each model wins and fails on my coding tasks.
by change size, checks to pass, repo size, how spread out the change is, and how models fail when they miss
workloadbench.brenhq.com/gpt…
1
217
Bren retweeted
how i split work across AI models
Opus 5.5 orchestrates my agents > Fable 5.2.
Fable 5.2 only for arch planning and research since a $200 Claude plan gets about 30 deep tasks a week of Fable.
We've given all Pro, Max, Ultra, and Team users one free full reset, across all models and limits. Use it anytime before December 10 at 2:00PM PT.
We recently fixed a few small issues that used more limits than intended, and the Fable resets didn't do much for those of you who'd already hit your weekly limits for all models. While these issues didn't affect everyone, we want to make up for the trouble.
You'll find it in Settings > Usage, or at clau.de/reset.
how i split work across AI models
Opus 5.5 orchestrates my agents > Fable 5.2.
Fable 5.2 only for arch planning and research since a $200 Claude plan gets about 30 deep tasks a week of Fable.
We've given all Pro, Max, Ultra, and Team users one free full reset, across all models and limits. Use it anytime before December 10 at 2:00PM PT.
We recently fixed a few small issues that used more limits than intended, and the Fable resets didn't do much for those of you who'd already hit your weekly limits for all models. While these issues didn't affect everyone, we want to make up for the trouble.
You'll find it in Settings > Usage, or at clau.de/reset.
GPT 6.1 Terra at max solved 20 of 30 tries on big changes.
xhigh solved 16, medium 3.
small fixes went from 85% at xhigh to 90% at max.
max costs $1.00 a run against $0.31 at xhigh.
Fable 5.2 streams 3x faster than GPT-6.1 SOL and finished my short code question no sooner
Introducing Claude Fable 5.2 and Claude Mythos 5.2, designed to tackle the most daunting technical challenges.
They’re capable collaborators on ambitious projects in many fields, and show improvements in areas like coding, reasoning, and vision.
1
222
Claude Fable 5.2 cost about twice what GPT-6 Astra did per run on my quick suite ticket test, $1.59 against $0.78
on bug hunt test they cost about the same, $1.18 against $1.06 to $1.21, and Fable 5.2 was quicker, 1.8 minutes against 2.3
Introducing Claude Fable 5.2 and Claude Mythos 5.2, designed to tackle the most daunting technical challenges.
They’re capable collaborators on ambitious projects in many fields, and show improvements in areas like coding, reasoning, and vision.
3
3
496
